Snake oil economics

The world’s biggest, richest, most comprehensive statistical agency has just announced significant revisions to the mainstream economists’ bellwether number – US GDP – and not just over the period of the last eight years, but for the last couple of quarters, too, we have seen a substantial alteration in the picture being painted (not that we have any grounds to suppose that the revisions, even now, are in any sense definitive).

Moreover, the most highly paid, highly vocal SELL side pundits who have been indoctrinated to rely on such a methodologically-suspect, empirically-unreliable aggregate of aggregates to help guide investment decision-making (oh, alright then: to generate commissions for their employers) are now scrambling pitifully to explain why they were wrong (yet again) in both their current analyses and their (typically over-optimistic) projections for what is to come – a plight made even worse by today’s ‘disappointment’ not just in the US purchasing managers’ report, but by the weaker numbers to be found in those emanating from elsewhere in the stimulus-addled Old World (flirting, in that narcotic’s absence with ‘triple dip’) and those erstwhile powerhouses of the recovery (now facing a looming stagflation) in the New one.

By contrast, it is a case of so far so good for our, Austrian-based ideas of enduring a marked deceleration over the course of the summer/fall, as first expressed a good six or seven months ago and reinforced frequently since. Does this contrast make us, personally, special?  No – but it surely underlines the fact that while there may be many, many economic analysts, there is only one correct way of performing economic analysis.

Returning to last Friday’s fiasco, let us note that, even if we were to set aside our deep-seated objections to the fatuous idea of trying to shoehorn the myriad activities of 300-odd million American men, women, and children – each unceasingly exchanging an uncountable variety of goods, services, and government ‘bads’, by means of a monetary circulation of some $100 billion a day –  into one single datum, the very, very best we can say of GDP is that it is akin to the frozen snapshot of reality provided by the earnings report of a vast, complex conglomerate, whose many, diverse, far-flung departments file incomplete reports, each compiled with differing degrees of diligence and rigour, and whose none-too scrupulous CFO has the ability greatly to recast the raw numbers within the wide latitude accorded him by the regulatory regime under which he operates.

Any one risking their capital in such a company without looking at the cash flow statement, the balance sheet, and the derivative footnotes – as well as taking a more objective view of the doings of that firm’s customers, suppliers, and competitors, alongside those of the broader market in which it operates – is leaving himself open to a severe loss of wealth.

Similarly, anyone analyzing an ‘economy’ – if there truly is such an animal – by fixating on this damnable number and/or its childishly tautologous, Keynes-Kuznetz building blocks (and especially, its narrow, final consumption constituents) without probing more deeply into the dynamics of the beast, is doomed before he starts. Any one who succumbs to the all-pervading temptation to appear authoritative by making scientific sounding extrapolations from blind, numerical entanglements of sub-indices whose own validity is equally questionable, is furthermore likely to suffer from a serial inability to make any demonstrably-skillful predictive statements about the shape of that uncertain future into which it is our intent honestly to try to peer.

Even if all that remains otherwise is to be confident of our deductive, microeconomic a prioris but only cautiously qualitative and broad brush in our prognoses of how all those individual choices may contribute to an incalculably complex whole, this is far and away preferable to the act of sheltering behind a façade of faux determinism. Remember that is unequivocally better – and certainly more useful – to be imprecisely right than to be embarrassingly wrong to three significant figures.

It is of course the case that so deeply engrained are statistics such as these in the vocabulary of both the market and the voodoo of Maynardian macro-economics that it is unrealistic to expect any practitioner to avoid any reference to them whatsoever. What is absolutely crucial, however, is (a) constantly to bear in mind that these are nothing but examples of a convenient shorthand which often conceal as much as they reveal – in the same way a mean height above sea level or an average annual temperature tells us little about the topography or climate of a region, much less about how those features may be changing – and (b) that the generation of a positive change in the metric is not an end in itself (as far too many policy jockeys and talking heads seem to believe). An amphetamine junkie getting his next fix by spending the contents of the old woman’s purse he just snatched generates more instant GDP than an engineer sitting quietly at his desk, trying to puzzle out a radical new way to create more useful output with less input, but it should be fairly obvious which man is likely to do more to improve both his own material comforts and those of the people around him.

Beware, then, the appallingly low signal:noise ratios in much of the data  – particularly when combining them arithmetically. The disproportionate (and typically unmentioned) impact on the small difference between two individually dubious estimates for the global  supply and demand for some key commodity, as these are continually cast and recast, is a classic case in point. Remember, too, that if an overly mechanical and mathematically-abstruse reasoning by induction from such data is the accepted way to build a glittering academic career – and even to head up a central bank or a supranational agency – it does very little to forewarn of the great turning points in gross economic activity, much less of short-term developments in the Greater Fool-populated financial markets which use the same numbers to help shape the landscape of the RPG universe which they inhabit.

If you follow this advice – humbling as it may be in terms of acknowledging your innate, human inability to foresee the details (rather than the rough direction) of the road ahead – you might just be able to give a good account of yourself the next time Her Britannic Majesty querulously asks you and your peers, “Why did nobody see it coming?”

More from Sean Corrigan
Beware of Mr. Wolf
The Saving Grace In a recent editorial, Martin Wolf of the Financial...
Read More
2 replies on “Snake oil economics”
  1. says: Paul Danon

    Please could this be rewritten so it’s readable? The sentences are a nightmare. Best regards.

  2. says: Tim Lucas

    A lovely turn of phrase:

    Remember that is unequivocally better – and certainly more useful – to be imprecisely right than to be embarrassingly wrong to three significant figures.

Comments are closed.